IDEAS AND TECHNICAL INNOVATIONS
Robot-assisted spine surgery: feasibility study througha prospective case-matched analysis
Nicolas Lonjon • Emilie Chan-Seng •
Vincent Costalat • Benoit Bonnafoux •
Matthieu Vassal • Julien Boetto
Received: 9 October 2014 / Revised: 2 January 2015 / Accepted: 2 January 2015
� Springer-Verlag Berlin Heidelberg 2015
Abstract
Purpose While image guidance and neuronavigation
have enabled a more accurate placement of pedicle
implants, they can inconvenience the surgeon. Robot-
assisted placement of pedicle screws appears to overcome
these disadvantages. However, recent data concerning the
superiority of currently available robots in assisting spinal
surgeons are conflicting. The aim of our study was to
evaluate the percentage of accurately placed pedicle
screws, inserted using a new robotic-guidance system.
Method 20 Patients were operated on successively by the
same surgeon using robotic assistance (ROSATM, Med-
tech) (Rosa group 10 patients, n = 40 screws) or by the
freehand conventional technique (Freehand group 10
patients, n = 50 screws). Patient characteristics as well as
the duration of the operation and of exposure to X rays
were recorded.
Results The mean age of patients in each group (RG and
FHG) was 63 years. Mean BMI and operating time among
the RG and FHG were, respectively, 26 and 27 kg/m2, and
187 and 119 min. Accurate placement of the implant (score
A and B of the Gertzbein Robbins classification) was
achieved in 97.3 % of patients in the RG (n = 36) and in
92 % of those in the FHG (n = 50). Four implants in the
RG were placed manually following failed robotic
assistance.
Conclusion We report a higher rate of precision with
robotic as compared to the FH technique. Provid-
ing assistance by permanently monitoring the patient’s
movements, this image-guided tool helps more accurately
pinpoint the pedicle entry point and control the trajectory.
Limitations of the study include its small sized and non-
randomized sample. Nevertheless, these preliminary results
are encouraging for the development of new robotic tech-
niques for spinal surgery.
Keywords Robot-assisted � Spine surgery � Lumbar �Degenerative disease
Introduction
Instrumented spinal surgery has seen massive growth over
recent years following the developmental boom in more
sophisticated spinal implants and modified surgical prac-
tice. Moreover, image-guided and neuronavigation tech-
niques have improved the accuracy of implant placement
and increased the safety of surgical procedures, particularly
those for which the radiation exposure has consequently
been reduced [1–5]. One recent meta-analysis performed
on clinical and cadaver studies reported an accuracy rate of
79 % (60–97.5 % SD 13 %) in the placement of lumbar
implants without use of a navigation system, and 96.1 %
(72–100 % SD 7.2 %) when navigation systems were used
N. Lonjon (&) � E. Chan-Seng � M. Vassal � J. Boetto
Department of Neurosurgery, Hopital Gui de Chauliac,
80 Avenue Augustin Fliche, 34090 Montpellier, France
e-mail: [email protected]
N. Lonjon
INSERM U1051, Institute for Neurosciences of Montpellier,
Pathophysiology and Therapy of Sensory and Motor Deficits,
Hopital Saint Eloi, Montpellier, France
V. Costalat
Department of Neuroradiology, Hopital Gui de Chauliac,
34090 Montpellier, France
B. Bonnafoux
Department of Clinic Research, Hopital la Colombiere,
39 Avenue Charles Flahault, 34295 Montpellier, France
123
Eur Spine J
DOI 10.1007/s00586-015-3758-8
[6]. However, the 15 studies on the lumbar spine included
in this meta-analysis used different criteria to assess
implant accuracy with some including pedicle breaches up
to 4 mm among their implants defined as accurately
placed. Clearly if we consider only intrapedicular implants
(respectively group A or 1 of Gertzbein and Robbins [7]
and Youkilis [8] classifications), as correctly positioned,
these rates sharply reduce and can reach 60 % for the
lumbar spine [9]. Today, no technique or device exists that
is able to guarantee 100 % accuracy in implant placement.
Robotic assistance in the placement of spinal implants is
however, showing great promise towards this ambitious
goal. The minimal invasive approach is perfectly suited to
this technique, because of the accuracy improvement and
the reduction of radiation exposure for both patient and
medical team. Yet, the technology developed thus far has
given mitigated results compared to conventional methods
with regards improving placement accuracy and reducing
radiation exposure. Thus, we carried out a prospective
feasibility study on pedicle screw implantation assisted by
the new robot ROSATM developed by the company Med-
tech. The results of 10 patients treated using this robotized
assistance were compared against those of 10 patients
consecutively treated by the conventional method.
Materials and methods
Two groups of patients undergoing spinal fusion for a
degenerative lumbar spine disease were prospectively
included without randomization. The first group (ROSA/
RG) comprised patients (n = 10) consecutively treated
using robotized assistance and the second group (Free-
Hand/FHG) comprised patients (n = 10) consecutively
treated via the conventional freehand approach during the
same period. The same surgeon operated all the patients
between April and October 2013. It was the first time this
surgeon used the robotic technology.
Patients
The patients included in this study fulfilled the following
criteria: (1) aged between 18 and 80 years; (2) indication of
posterior fusion for a degenerative lumbar spine disease
(lumbar stenosis, degenerative disc disease, degenerative
lumbar spondylolisthesis); (3) signed informed consent to
the use of robotized assistance and the planned surgical
procedure.
Exclusion criteria were: (1) previous surgery on the
lumbar spine; (2) need for lumbosacral posterior fusion.
Due to the differences in anatomy and trajectory plan-
ning on 2D fluoroscopic images between S1 and other
lumbar vertebrae, we excluded all patient requiring a spinal
fusion with S1 pedicle screw insertion to avoid bias in our
comparison between the two groups of both groups.
As this represented a pilot feasibility study on the use of
a new robotized assistance system, all procedures were
performed by a conventional open posterior median
approach. The intraoperative visualization of anatomical
landmarks thus enabled a view of and control over the
trajectories proposed by the robotized assistance.
The surgical indications of osteosynthesis, laminectomy
and interbody fusion (cage) were left to the surgeon’s
discretion in accordance with the underlying disease and in
respect of best current practice.
The local ethics committee for the protection of humans
(Comite de Protection des personnes) and the national drug
safety agency ANSM (Agence Nationale de Securite du
Medicament et des produits de sante) approved the study.
Conventional osteosynthesis: freehand group (FHG)
A mobile C-arm with image intensifier was positioned in
the lateral plane during the conventional freehand proce-
dure. The pedicle screws were implanted using anatomical
landmarks and by progressive pedicular palpation. Anter-
oposterior and lateral fluoroscopic control was performed
at the end of the procedure. A laminectomy and/or inter-
somatic fusion (cage TLIFT) were performed if necessary
according to the patient’s disease.
Robot-assisted osteosynthesis: ROSA group (RG)
Robot-assisted osteosynthesis performed on RG patients
used the new robot ROSATM developed by Medtech. This
device comprises a first floor-fixable mobile base onto
which is mounted a six axes robot arm, and a second
mobile base onto which is mounted a navigation camera.
The robot arm supports a tool that helps to guide conven-
tional neurosurgical instruments, and allows their accurate
and stable positioning onto the target point planned by the
surgeon. The surgeon remains in complete control of the
surgical act while benefiting from the accuracy and sta-
bility offered by the robotized assistance in identifying the
pedicle entry point and controlling the trajectory in all
circumstances. Indeed, all patient’s movements due either
to patient respiration or to acts executed by the surgeon are
monitored by the navigation camera with the help of a
patient reference implanted in the iliac crest. The device
adjusts in real time the robot position to all spine move-
ments. Thus, the tool supported by the robot remains
always aligned on the entry point and the orientation
defined by the surgeon during the planning.
ROSATM is an image-guided device, which positions the
tool to be guided according to landmarks defined either
directly on intraoperative radiographic images or using a
Eur Spine J
123
navigation pointer. The guidance is based on trajectory
planning performed with the intraoperative 2D image treat-
ment software, followed by the registration of the patient in
prone position (Fig. 1). At the beginning of the procedure,
the robot is brought along the right-hand side of the patient,
ensuring that the robotized arm is able to sufficiently cover
the two levels spinal segment concerned (Fig. 2). The sur-
geon then stands at the other side of the patient. The oper-
ating room is organized as represented in Fig. 3.
ROSATM Spine is an image-guided device, which
combines robotic assistance in positioning tools according
to planned trajectories and navigation features.
Once the patient is in surgical position, the ROSA platform
is installed with the robot arm stand on the side of the operating
table and the navigation camera stand on the patient foot side.
All patient’s movements are followed with the help of a spe-
cific patient reference target attached to the iliac crest.
An acquisition of intraoperative fluoroscopic x-rays is
performed using a registration pattern held by the robot arm.
On a specially designed graphic user interface with proprie-
tary software, the surgeon plans the trajectory of the screws.
A tool holder guide is positioned at the extremity of the
articulated robotic arm. Said tool holder acts as a axial
physical guide for most conventional neurosurgical
instruments. The navigation software coupled with the
axial guide enables accurate execution of the planned tra-
jectories. It leaves the surgeon all his tactile feedback
during tool insertion in bones.
The system allows for real-time adjustment of the robot
trajectories thanks, to a permanent monitoring of the
patient’s movements due either to patient breathing or to
acts executed by the surgeon.
The platform has been designed to cover a high number
of vertebrae, without having to move it between each
levels. For this feasibility study, we voluntarily limited the
scope to two lumbar levels.
Studied parameters
The main aim of the study was to measure and compare the
placement of pedicle screws by the two surgical techniques
described above. Two investigators studied the postopera-
tive CT scans of the spine with coronal and sagittal image
reconstruction to evaluate each implantation according to the
classification of Gertzbein and Robbins [7] and a modified
classification of Youkilis [8] (Fig. 4). In cases where the two
investigators disagreed on implant position, a further CT-
scan review enabled consensus to be reached. The two
investigators, an independent neurosurgeon (V. M.) and an
independent radiologist (C. V.), were blinded to the tech-
nique used.
For each RG implant, every attempt was described as:
(1) screw implantation successful via ROSATM robotic
assistance; (2) attempt interrupted due to technical failure
of robotic assistance and implant placed manually; or (3)
Fig. 1 Preoperative fluoroscopy-based planning of the transpedicular screws
Eur Spine J
123
trajectory and/or implantation failed by ROSATM robotic
assistance and implant repositioned manually.
The secondary objectives concerned the duration of the
surgery, time in the operating room, total dose of radiation,
total duration of radiation exposure, and the number of
radiographic images taken for each patient.
The characteristics age, sex, BMI and length of hospi-
talization were recorded for each patient.
The level of instrumentation, the number of screws
implanted, and whether a laminectomy had been performed
or an intersomatic cage been placed, were noted for each
procedure.
Statistical analysis
Continuous variables were compared using the Mann–
Whitney rank-sum test. The Chi-square test or the Fisher
exact test was used to compare categorical variables.
Results
For a period of 5 months, between April and October 2013,
we non-randomly included 20 patients in the study. Patient
characteristics and details of the procedure performed are
Fig. 2 Preoperative view, with an overall view of the robot-assisted: ROSA (a), and the surgical tool holder view (b) and surgical view with the
working robot (c)
Fig. 3 Operative room layout
Eur Spine J
123
Fig. 4 Method of screw assessment according to Gertzbein and
Robbins classification with an example of grade 1 (a), grade 2 (b) and
grade 4 (c). a Example patient n�7 (left L4 screw)-grade 1 of the
Gertzbein and Robbins classification. b Example patient n�3 (left L4
screw)-grade 2 of the G. and R classification. c Example patient n�5
(Right L4 screw)-grade 4 of the G. and R classification
Eur Spine J
123
summarized in Table 1. The groups displayed no signifi-
cant differences with regards median age of the patients,
sex ratio, median BMI and proportion of laminectomies
and intersomatic fusions performed. The distribution
according to the level of osteosynthesis is represented in
Table 2.
Both operating room time (ORT) and surgery time (ST)
were much longer among the RG patients (respectively,
336 ± 20 and 186 ± 21 min) compared to the FHG
patients (209 ± 29 and 112 ± 29 min), with RG patients
spending approximately 2 h more in the operating room
and 1 h more undergoing the surgical procedure.
Radiation exposure expressed as the Dose Area Product
(absorbed dose in cG multiplied by area irradiated in cm2)
was 821 ± 585 and 406 ± 315 cGy cm2 for the RG and
FHG, respectively. Duration of fluoroscopy was respec-
tively 1.23 ± 0.37 and 0.40 ± 0.28 min in the RG and
FHG implying that the robotic assistance doubled the
radiation exposure time of patients. A total number of 53
radiographies with an average of 3.5 per patient were
required in the RG compared to none in the FHG.
The position of the implants according to Gertzbein and
Robbins classification and a modified Youkilis classifica-
tion (grade 4 was separated in two to specify the nature of
the pedicle breach, 4A medial breach; 4B lateral breach) is
represented in Table 3. By considering only implants
placed under robotic assistance for the RG, the rate of
accurate implant positioning (grade A or B of Gertz. and R
classification, or grade 1 or 2 of Youkilis classification)
was 97.3 % in the RG compared to 92 % in the FHG. The
results and p values are detailed in Table 3. On three
occasions during screw insertion, the robotic assistance in
RG patients was interrupted due to a loss of signal (2
screws) during the procedure or computer crash due to
software failure (1 screw). Fluoroscopic control detected
only one implant that had been incorrectly positioned under
robotic assistance, which had to be re-positioned manually.
The patient in this case presented a slight degenerative
scoliosis and the lateral projection of the two vertebrae
considered for fusion was not ideal. The implant, which
had been positioned on a too ascending trajectory and had
burst into the intersomatic space, was thus re-positioned.
But none of the misplaced screws with pedicle breach led
to neurologic consequences, or other complications for the
patients from both groups.
Discussion
In their meta-analysis, Kosmopoulos et al. reported an
average increase of 5 % in the accuracy rate of pedicle
implant placement using navigation systems without robotic
assistance. In their study using the O-Arm Multidimensional
Surgical Imaging System to place percutaneous lumbar
screws, Houten et al. [10] found a high rate of accuracy with
97 % classified as grade A. However, in less well-trained
teams, results with use of computer-assisted navigation
Table 1 Study parameters
FHG (n = 10) RG (n = 10) P value
Median: average age: year ± SD (range) 63: 63.4 ± 11 (45–79) 60.5: 63.4 ± 11 (50-79) 0.860
Sex (male/female) 4/6 4/6 0.999
Median: average BMI ± SD (range) 26.4: 27.3 ± 5.6 (18–36) 26.7: 27.8 ± 4 (23–35) 0.809
Laminectomy 9/10 8/10 0.999
Intersomatic fusion (cage: TLIF) 4/10 5/10 0.670
Median ORT: min ± SD (range) 209 ± 29 (167–270) 336 ± 20 (318–375) 10-4
Median ST: min ± SD (range) 112 ± 29 (76–192) 186 ± 21 (150–213) 10-4
Median DAP: cGy.cm2 ± SD (range) 406.79 ± 315 (355–1,186) 821.63 ± 585 (654–2,195) 10-4
Median RET min (range) 0.40 ± 0.28 (0.03–1.27) 1.23 ± 0.37 (1.12–2.43�) 0.008
Radiographies number: (total nb/nb median per patient) 0 53/3.5 \10-16
Average hospital stay (days) 6.9 (5–10) 6.67 (5–9) –
SD standard deviation, FH free hand, BMI body mass index, TLIF transforaminal lumbar interbody fusion, ORT operating room time, ST surgery
time, DAP dose area product, RET radiation exposure time, nb number, FHG freehand group, RG ROSA robotic assistance group
Table 2 Number of screws per level and group
Instrumented VB FHG (n = 50) RG (n = 40) Total
L2 6 0 8
L3 8 6 14
L4 20 20 40
L5 16 16 32
VB vertebral body, FHG freehand group, ROSA robotic assistance
group
Eur Spine J
123
systems in percutaneous surgery are much lower, with
sometimes up to 15 % pedicle breach, thus requiring the
development and use of robotic assistance [11, 12].
Image-guided robotic assistance theoretically allows the
increase in accuracy notably in percutaneous indications,
by restricting to the planned trajectory on 2D or 3D images
and above all by adapting to intraoperative movements of
the patient.
In practice, however, use of robotic assistance has thus
far given conflicting results: one retrospective study on 646
screws implanted by robotic assistance evaluated by post-
operative CT scan and Gertzbein and Robbins classifica-
tion found that 98.3 % had been accurately positioned
(grade A or B) [13]. In their prospective and randomized
study, Roser et al. [14] reported a rate of 99 % accurate
placement (grade A only) with robotic assistance provided
by SpineAssist from Mazor Surgical technologies, as
compared to 97.5 % using the freehand technique and
92 % with navigation. In contrast, Ringel et al. [15] dem-
onstrated an inferior performance among the robotic-
assisted group (SpineAssist) in their prospective and ran-
domized study, with only 85 % meeting the grade A–B
criteria versus 93 % in the freehand conventional group.
The method of fixation of the miniature robot on the patient
and deviation of a too flexible cannula upon contact with
the articular surface could explain these differences [15].
Here we have evaluated the accuracy of a new technique
of image-guided robotic assistance developed for spine
Table 3 Screw position
according to the Gertzbein and
Robbins and Youkilis
classification, given in total
number and percentage of
implanted screws
Gertz. and R. Gertzbein and
Robbins, nb number, N/A not
applicable
Table 3
Screw position according to the Gertzbein and Robbins and Youkilis classification, given in total
number and percentage of implanted screws
Screw position FHG (n=50)
nb/%
GR (n=40)
nb/%
Total (n=90)
nb/% P valueGertz. & R. Youkilis
A 1 42 / 50 (84%) 33 / 36 (89.2%) 75/86 (87.2%) -
B 2 4 2 6 -
A+B 1+2 46 / 50 (92%) 35 / 36 (97.3%) 81/86 (94.2%) 0.639
C 3 1 1 0 0 2 -
D 4A 2 1 0 1 2 1 -
E 4B 1 2 1 0 1 2 -
C+D+E 3+4A+4B 4/50 (8%) 1/36 (2.7%) 5/86 (5.8%) 0.639
Successful procedure with
robotic guidanceN/A 36/40 (90%) N/A N/A
Use of robot aborted N/A 3/40 (7.5%) N/A N/A
Immediate manual
conversionN/A 1/40 (2.5%) N/A N/A
Gertz. & R. : Gertzbein and Robbins; nb: number; N/A: Not applicable
Eur Spine J
123
surgery. Our pilot study was in effect to validate a new
application of the ROSATM robot that was originally ded-
icated to cranial neurosurgery. Thus, to ensure maximum
safety, all patients underwent open surgery allowing ana-
tomic control over the pedicle entry points. The accuracy
rate of 92 % (Gertzbein and Robbins classification) that we
achieved in implant placement via the conventional free-
hand technique agrees with data in the literature on the
lumbar segment [6, 15]. We also found a greater accuracy
in implant placement offered by robotic assistance as
compared to the conventional freehand technique (97.3 %
of grade A–B versus 92 %).
The design of the ROSATM robotic assistance, with its
compact floor-fixable base and rigid robotized arm, reduces
the risk of secondary movement of the equipment and
guarantees the reliability of the entry point and trajectory.
The method of permanently monitoring patient’s move-
ments further reinforces this reliability. The poor posi-
tioning of implants (grade C or D) in our series (3/36,
2.7 % of the implants), was likely in large part due to poor
planning of the trajectory on only one anterior-posterior
and one lateral X-ray intraoperative view and could be
considerably reduced with use of 3D images.
This could partly be explained by the difficulty in
imposing the entry point on dense cortical bone when the
theoretical trajectory of the implant is not tangent to the
surface at the penetration point (curved articular surface).
Sliding of the tool as well as a change in trajectory during
guide wire insertion into the bone can result from lack of
rigidity of the cannula and/or of the wire. These findings
were also reported by Ringel et al. [15] and can be avoided
in future procedures with simple technical modifications.
Concerning radiation exposure, we did find a more than
twofold increase in both exposure time and Dose Area
Product among the group operated on by robotic assistance
compared to that by the conventional technique. This can
partly be explained by the need to perform radiography,
which is more radiating than fluoroscopy to improve the
quality of the X-ray image onto which the trajectory of the
implant is planned. In theory, only two images (AP and
lateral) are required for image matching (registration)
between patient and robot and yet an average of 5.3 images
was taken per patient in our study, notably to perform
virtual superimposition analyses on the image of the screws
implanted with those planned. This step allowed us to
verify the overall reliability of robotic assitance. Also,
fluoroscopic images were performed at the surgeon’s dis-
cretion to definitively validate the implant positioning. This
latter step in addition to a number of the fluoroscopic
images taken during the intervention to check each implant
will not be necessary in the future.
The radiation exposure time did however, remains rea-
sonable: 25 s/screw in the RG group and 10 s/screw in the
FHG, against 20 s/screw via a conventional surgery tech-
nique in another study on 140 patients [16]. In other studies
using robotized assistance, Kantelhardt et al. [17] reported
an average exposure time of 77 s/screw by conventional
surgical method and 44 s/screw using robotic assistance in
open surgery and Roser et al. [14] reported, respectively,
31.5 and 16 s.
The increase in both radiation exposure and surgical
time can be partly attributed to the learning curve and the
setting of the feasibility study, as well as the setup and the
prototype nature of the first patient series.
Our study has certain limitations, the most important
being the small number of patients and the absence of
randomization. Also, the method the most used namely,
defining the entry point using both visual anatomical
landmarks and AP and lateral radiographic landmarks
(navigation pointer) simultaneously, without doubt
strengthened the accuracy of implant positioning in our
series. While the results presented here do favor the better
accuracy achieved with robotic assistance, the data should
be confirmed on a larger sample and on percutaneous
procedures.
Conclusion
The preliminary results of this study are encouraging. The
number of patients is small, yet the feasibility of use and
the reliability of this new robotic assistance have been
demonstrated. On the series of patients in our study, this
robotic tool improved the accuracy of pedicle implant
placement and ensured an optimized safety of the surgical
procedure. The fact that the system used in this study is still
under development partly explains the higher radiation
exposure, and the longer surgery and operating room times
found in the group undergoing robotic-assisted surgery as
compared to that operated on using the conventional free-
hand technique. Future studies on percutaneous surgery are
planned to definitively validate this tool. Continuing
improvements in robotics technology over the coming
years promises their application in spine surgery and fur-
ther improved patient safety.
Conflict of interest We declare a consulting agreement between
Medtech and the first author Nicolas Lonjon.
References
1. Kosmopoulos V, Theumann N, Binaghi S, Schizas C (2007)
Observer reliability in evaluating pedicle screw placement using
computed tomography. Int Orthop 31:531–536
2. Merloz P, Tonetti J, Cinquin P, Lavallee S, Troccaz J, Pittet L
(1998) Computer-assisted surgery: automated screw placement in
the vertebral pedicle. Chir Mem Academie Chir 123:482–490
Eur Spine J
123
3. Tian N-F, Huang Q-S, Zhou P, Zhou Y, Wu R-K, Lou Y, Xu H-Z
(2011) Pedicle screw insertion accuracy with different assisted
methods: a systematic review and meta-analysis of comparative
studies. Eur Spine J 20:846–859
4. Van de Kelft E, Costa F, Van der Planken D, Schils F (2012) A
prospective multicenter registry on the accuracy of pedicle screw
placement in the thoracic, lumbar, and sacral levels with the use
of the O-arm imaging system and StealthStation Navigation.
Spine 37:E1580–E1587
5. Verma R, Krishan S, Haendlmayer K, Mohsen A (2010) Func-
tional outcome of computer-assisted spinal pedicle screw place-
ment: a systematic review and meta-analysis of 23 studies
including 5,992 pedicle screws. Eur Spine J 19:370–375
6. Kosmopoulos V, Schizas C (2007) Pedicle screw placement
accuracy: a meta-analysis. Spine 32:E111–E120
7. Gertzbein SD, Robbins SE (1990) Accuracy of pedicular screw
placement in vivo. Spine 15:11–14
8. Youkilis AS, Quint DJ, McGillicuddy JE, Papadopoulos SM
(2001) Stereotactic navigation for placement of pedicle screws in
the thoracic spine. Neurosurgery 48:771–778 (Discussion
778–779)
9. Castro WH, Halm H, Jerosch J, Malms J, Steinbeck J, Blasius S
(1996) Accuracy of pedicle screw placement in lumbar vertebrae.
Spine 21:1320–1324
10. Houten JK, Nasser R, Baxi N (2012) Clinical assessment of
percutaneous lumbar pedicle screw placement using the O-arm
multidimensional surgical imaging system. Neurosurgery
70:990–995
11. Nakashima H, Sato K, Ando T, Inoh H, Nakamura H (2009)
Comparison of the percutaneous screw placement precision of
isocentric C-arm 3-dimensional fluoroscopy-navigated pedicle
screw implantation and conventional fluoroscopy method with
minimally invasive surgery. J Spinal Disord Tech 22:468–472
12. Von Jako R, Finn MA, Yonemura KS, Araghi A, Khoo LT,
Carrino JA, Perez-Cruet M (2011) Minimally invasive percuta-
neous transpedicular screw fixation: increased accuracy and
reduced radiation exposure by means of a novel electromagnetic
navigation system. Acta Neurochir (Wien) 153:589–596
13. Devito DP, Kaplan L, Dietl R, Pfeiffer M, Horne D, Silberstein B,
Hardenbrook M, Kiriyanthan G, Barzilay Y, Bruskin A, Sackerer
D, Alexandrovsky V, Stuer C, Burger R, Maeurer J, Donald GD,
Gordon DG, Schoenmayr R, Friedlander A, Knoller N, Schmie-
der K, Pechlivanis I, Kim I-S, Meyer B, Shoham M (2010)
Clinical acceptance and accuracy assessment of spinal implants
guided with Spine assist surgical robot: retrospective study. Spine
35:2109–2115
14. Roser F, Tatagiba M, Maier G (2013) Spinal robotics: current
applications and future perspectives. Neurosurgery 72(Suppl
1):12–18
15. Ringel F, Stuer C, Reinke A, Preuss A, Behr M, Auer F, Stoffel
M, Meyer B (2012) Accuracy of robot-assisted placement of
lumbar and sacral pedicle screws: a prospective randomized
comparison to conventional freehand screw implantation. Spine
37:E496–E501
16. Jones DP, Robertson PA, Lunt B, Jackson SA (2000) Radiation
exposure during fluoroscopically assisted pedicle screw insertion
in the lumbar spine. Spine 25:1538–1541
17. Kantelhardt SR, Martinez R, Baerwinkel S, Burger R, Giese A,
Rohde V (2011) Perioperative course and accuracy of screw
positioning in conventional, open robotic-guided and percutane-
ous robotic-guided, pedicle screw placement. Eur Spine J
20:860–868
Eur Spine J
123